The hypothesis of continuity between childhood-onset and adult schizophrenia was tested by comparing the performance of 15 patients with childhood-onset schizophrenia and 52 age-matched controls on 2 reaction time paradigms that have been used to study adult schizophrenia. On simple reaction time to tones with regular and irregular preparatory intervals of 2, 4, and 8 s, patients showed greater effects of the length of the preparatory interval in the regular condition and greater effects of the preparatory interval (girls only) and the preceding preparatory interval in the irregular series. On simple reaction time to random lights and tones, patients were faster on ipsimodal sequences than cross-modal sequences compared with controls. Overall, patients were much slower than controls in both paradigms. The results suggest similar attention dysfunction as is found in adult schizophrenia and thus are consistent with the continuity hypothesis. Symptomatic schizophrenia typically develops in late adolescence to early adulthood, but some apparent cases have earlier onset, extending into the preadolescent period (Karno & Norquist, 1989; Kramer, 1978). There has been some controversy about whether the childhood-onset disorder (defined here as having an onset at age 12 years or younger) is fundamentally the same disorder as the more common adult variety. Although child schizophrenia was once classified with infantile autism as a distinct class of disorders (American Psychiatric Association, 1968; Bender & Faretra, 1972), careful studies of the phenomenology and clinical course of the disorder have indicated continuity between childhood-and adult-onset forms of the illness (